Layer 30 · Neuron 1742 · suppressed

L30N1742-1

Word quirks:

“围”“狱”“魄”“ประท”“ikes”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Qwen3.5, a large-scale language model developed by Tongyi Lab. My core mission is to assist users in solving tasks across a wide range of domains—such as writing, coding, data analysis, and logical reasoning—while ensuring safety, accuracy, and efficiency. I aim to be a versatile, reliable, and human-like assistant that adapts to diverse needs, from casual conversations to complex problem-solving scenarios. Whether you need help creating content, analyzing data, or exploring new ideas, I'm here to support you with precision and care. How can I assist you today?

Nearest neighbors

Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.

By word quirks Profile recipe

Specimen	Similarity	Word quirks
1. L31N5934-1💥	0.79	“瑞”“虏”“宇”“虑”“饰”
2. L26N8298-1💤	0.78	“仕”“商”“坚毅”“皮”“朝”
3. L31N4653+1💥	0.78	“院”“顾”“轮”“grab”“核”
4. L28N3671+1💤	0.78	“希”“御”“沙”“璋”“乐”
5. L31N1613-1💥	0.77	“滤”“嫦”“爛”“拷”“刽”
6. L29N7374-1🕒	0.77	“节”“炎”“气”“营”“斑”
7. L29N5128+1🕒	0.77	“福”“雷”“布”“波”“法”
8. L26N7921+1💤	0.77	“认”“模”“垂”“RP”“佩”
9. L28N4632+1💤	0.77	“司”“泡”“反”“路易”“收”
10. L29N7586+1🕒	0.77	“odia”“圈的”“一出”“周”“圈”

Shadow specimen, apart from the ten: L30N1742+1💤, 0.12.

Cousins · specimens sharing a neuron

No specimen with replies shares this neuron yet.

Technical details

Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.

Layer 30, neuron 1742, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.25, top-1 agreement 0.82) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. See Method for the tiers.

Quick scan · prediction measurements

Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.

KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.

The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.

Setting	KL divergence	Top-1 agreement	ΔNLL	Entropy	Steer lens
L30N1742+1💤 shadow specimen t0-v0.1	0.309	0.83	0.377	1.43	“وام”“vah”“эк”“esti”“_FAILURE”
L30N1742-1 t0-v0.1	0.254	0.82	0.331	1.58	Top boosted tokens ↑

Method

A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.

Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.